99 Problems But Matching Algorithms Ain’t One
Every site from eHarmoney to OKcupid loves to tout about their “secret sauce” matching algorithms. We’re sure you have seen the cheesy commercials boasting about how their sophisticated matching systems. We’ve done a lot of research here at Hey There and we asked ourselves…
How they (Match.com, OKcupid, eHarmony and others) can find singles everywhere a uniquely compatible soul mate at a click of a button is a mystery to all of us. Or is it? We are here to tell you it’s all FAKE NEWS! There is no scientific data to support any claim of matching algorithm superiority. Any claim is negligible at best!
Without any empirical data to support claims of matching algorithm success how can anyone say with any certainty that they are superior to any other form of matching or even matching people randomly? How do you honestly know their “algorithms” are actually taking into account any data at all and aren’t just literally “matching” you to whoever pops up in their system? Because they say so? really? You take them at their word?
Until the scientific community can get their hands on these mathematical cupid formulas to evaluate we are left to question their legitimacy. And there’s plenty of online daters who are NOT getting good matches who inevitably lead to question exactly that. Although we cannot produce a shred of evidence to back up their success due to their unwillingness to publish their algorithms we can make a determination about their fundamental flaw.
Mathematical matching algorithms are powered by two principles: similarities and opposite qualities. It is pretty much common sense to think that a match is best made when two people have more similarities versus people that are more dissimilar. It is also fair to go by the old adage that “opposites attract.” But don’t those two contradict each other? Even though it might make sense superficially, there is little evidence to support it as scientifically true. In fact there have been scientific studies that suggest neither principle has any impact on relationship quality. There are too many factors at play in human relationship building to think for a single second that a cookie cutter algorithm could predict anything. Far too many!
The facts are that no matching algorithm can predict long term success. They lack real life interactions and social circumstances. The algorithms are only as good as the data they process and so far the data is circumstantial at best. We have to realize that a lot of times the answers given to the questionnaires on dating platforms are often falsified and/or bent truths. Real life insights are the data points that matter. But how can we get that online?
Let me introduce you to “Hey There” which was created because it’s time for a change of approach.
Our theory is that, even though we can’t process interaction between two people to predict relationship success we can at least provide real world insights. That is why we base initial matching off basic data like location, age, gender, and interests. But then we combine that with social insights from friends that have true relevance to matching. We believe where math matchmaking falls short our system based on people matching works best. Insights come from us as humans not machines.